Thinking Skills and Creativity, Journal Year: 2024, Volume and Issue: 51, P. 101474 - 101474
Published: Feb. 6, 2024
Language: Английский
Nature Human Behaviour, Journal Year: 2024, Volume and Issue: 8(6), P. 1044 - 1052
Published: May 13, 2024
Language: Английский
Citations
52
ACM Computing Surveys, Journal Year: 2024, Volume and Issue: 56(11), P. 1 - 36
Published: June 25, 2024
Misinformation is one of the key challenges facing society today. User-centered misinformation interventions as digital countermeasures that exert a direct influence on users represent promising means to deal with large amounts information available. While an extensive body research this topic exists, researchers are confronted diverse landscape spanning multiple disciplines. This review systematizes user-centered facilitate knowledge transfer, identify trends, and enable informed decision-making. Over 6,000 scholarly publications were screened, systematic literature ( N=172 ) was conducted. A taxonomy derived regarding intervention design (e.g., labels, showing indicators misinformation, corrections, removal, or visibility reduction content), user interaction (active passive), timing pre post exposure request user). We provide structured overview approaches across disciplines derive six overarching for future transferability (1) novel platforms (2) emerging video- image-based sensible combination automated mechanisms (3) human experts (4) feedback comprehensibility, (5) encouraging media literacy without exposure, (6) adequately addressing particularly vulnerable such older people adolescents.
Language: Английский
Citations
18
Frontiers in Artificial Intelligence, Journal Year: 2025, Volume and Issue: 8
Published: April 7, 2025
The emergence of artificial intelligence (AI) large language models (LLMs), which can produce text that closely resembles human-written content, presents both opportunities and risks. While these developments offer significant for improving communication, such as in health-related crisis they also pose substantial risks by facilitating the creation convincing fake news disinformation. widespread dissemination AI-generated disinformation adds complexity to existing challenges ongoing infodemic, significantly affecting public health stability democratic institutions. Prompt engineering is a technique involves specific queries given LLMs. It has emerged strategy guide LLMs generating desired outputs. Recent research shows output depends on emotional framing within prompts, suggesting incorporating cues into prompts could influence their response behavior. In this study, we investigated how politeness or impoliteness affects frequency generation various We generated evaluated corpus 19,800 social media posts topics assess capabilities OpenAI's LLMs, including davinci-002, davinci-003, gpt-3.5-turbo, gpt-4. Our findings revealed all efficiently (davinci-002, 67%; 86%; 77%; gpt-4, 99%). Introducing polite prompt requests yielded higher success rates 79%; 90%; 94%; 100%). Impolite prompting resulted decrease production across 59%; 44%; 28%) slight reduction gpt-4 (94%). study reveals tested effectively generate Notably, had impact rates, with showing when prompted compared neutral impolite requests. investigation highlights be exploited create emphasizes critical need ethics-by-design approaches developing AI technologies. maintain identifying ways mitigate exploitation through crucial prevent misuse purposes detrimental society.
Language: Английский
Citations
2
Published: Dec. 16, 2022
The spread of misinformation through media and social networks threatens many aspects society, including public health the state democracies. One approach to mitigating impact focuses on individual-level interventions, equipping policy-makers with essential tools curb influence falsehoods. Here we introduce a toolbox individual-focused interventions aimed at reducing harm from online misinformation. Comprising an up-to-date account featured in 81 scientific papers across globe, is resource for scientists, policymakers, public. It provides both conceptual overview breadth interventions---including their target, scope, examples---and summary empirical evidence supporting methods experimental paradigms used test them. covers nine categories interventions: accuracy prompts, debunking rebuttals, friction, inoculation, lateral reading verification strategies, media-literacy tips, norms, source-credibility labels, warning fact-checking labels.
Language: Английский
Citations
48
Nature Human Behaviour, Journal Year: 2024, Volume and Issue: 8(5), P. 823 - 834
Published: March 18, 2024
Language: Английский
Citations
12
Published: May 30, 2022
Efforts to combat misinformation have intensified in recent years. In parallel, our scientific understanding of and information ecosystem has improved. Here, I propose ways improve interventions against based on this growing body knowledge. First, because consumption is minimal news low, more should aim at increasing the uptake reliable information. Second, most people distrust unreliable sources but fail sufficiently trust sources, there room than reduce sources. Third, largely a symptom deeper socio-political problems, try address these root causes, such as by reducing partisan animosity. Fourth, small number powerful individuals give its visibility, target ‘superspreaders’. Fifth, false not necessarily harmful true can be used misleading ways, misleadingness take precedence over veracity defining misinformation. Policymakers, journalists, researchers would benefit from considering arguments when thinking about problem how tackle it.
Language: Английский
Citations
30
European Psychologist, Journal Year: 2023, Volume and Issue: 28(3), P. 162 - 172
Published: July 1, 2023
Abstract: What does health misinformation look like, and what is its impact? We conducted a systematic review of 45 articles containing 64 randomized controlled trials (RCTs; N = 37,552) on the impact behaviors their psychological antecedents. applied planetary perspective by framing environmental issues as human focusing about diseases, vaccination, medication, nutrition, tobacco consumption, climate change. found that in 49% cases exposure to damaged antecedents such knowledge, attitudes, or behavioral intentions. No RCTs evaluated direct measures pro-environmental (e.g., vaccination), few studies explored feelings, social norms, trust. Most was based logical fallacies, conspiracy theories, fake experts. evaluating impossible expectations cherry-picking are scarce. research focused healthy adult US populations used online samples. Future can build our analysis address knowledge gaps we identified.
Language: Английский
Citations
18
International Journal of Human-Computer Interaction, Journal Year: 2024, Volume and Issue: unknown, P. 1 - 28
Published: Jan. 17, 2024
Interest in technological solutions for combating online misinformation has rapidly grown over the last decade, yet majority of proposed tools do not consider behavioral theories their design, nor have they addressed ways which individuals could potentially interact with these tools, while omitting plausible variations design interventions may affect end-users' decision-making and responses. In this paper, we explore potential nudging to inform that aim mitigate through behavior change. We report on a workshop where 29 participants were asked conceive technology-mediated nudges supporting individuals' production, dissemination or consumption misinformative content. producing novel solutions, used "Nudge Deck," support tool makes nudge theory, particularly framework 23 interaction mechanisms nudging, accessible during time-constrained meetings. present outputs session discuss them light prior literature respect ethics effectiveness.
Language: Английский
Citations
6
European Journal of Public Health, Journal Year: 2024, Volume and Issue: 34(2), P. 380 - 386
Published: Feb. 20, 2024
Abstract Background Predictors of COVID-19 (coronavirus) vaccination have been extensively researched; however, the contextual factors contributing to understanding intention remain largely unexplored. The present study aimed investigate moderating role economic development (Gross domestic product - GDP per capita), inequality (Gini index), perceived corruption index and Hofstede’s measurements cultural values—index individualism/collectivism power distance index—in relationship between determinants satisfaction with healthcare system, trust in political institutions, conspiracy beliefs intention. Methods A multilevel modelling approach was employed on a sample approximately 51 000 individuals nested within 26 countries. Data were drawn from European Social Survey Round 10. model examined effect individual- country-level predictors their interaction Results Satisfaction system had stronger positive get vaccinated countries lower more individualistic Trust institutions higher corruption, while negative development, collectivistic orientation. Conclusion Our findings highlight importance considering individual when addressing
Language: Английский
Citations
5
Journal of Research on Technology in Education, Journal Year: 2024, Volume and Issue: unknown, P. 1 - 27
Published: April 19, 2024
Although the serious game Bad News has been used to inoculate citizens against misinformation, it not formally evaluated in traditional classrooms. We therefore its impact on 516 upper-secondary Swedish students playing individually, paired, or with whole class. Results show that improved their ability discern manipulation techniques social media posts. Students prior positive attitudes credible news sources were better discerners, and this attitude became significantly more post-intervention. Rationales for identifying manipulative increased among those who credibility ratings pre-to Lastly, enjoyment of interest intervention was higher whole-class setting. This study offers insights educators using games formal teaching foster information literacy.
Language: Английский
Citations
5